The growing number of services that run machine learning algorithms under the hood will now be able to tap the popular open-source database Redis’ capabilities much more easily than before.

Typically used to cache applications’ most frequently accessed information, Redis has received a new extension today that brings integration with Spark ML. Organizations that employ that framework to power their machine learning projects can now easily load models into Redis to improve their performance. Initial benchmark tests indicate that the system provides five to 10 times lower latency than a standard Spark implementation when performing certain tasks such as categorizing records.

The assessment was carried out by Redis Labs Inc., the startup behind the new module and the database’s main commercializer. Besides improving response times, the firm also claims that today’s integration has the potential to streamline application development a great deal. The reason is that machine learning models kept on the system can be accessed using a wide range of languages and not just the native syntax in which they are written.

As a result, engineers don’t have to incorporate any extra wrapper libraries into their projects to provide interoperability. And the amount of code required to deploy a Spark ML model is further reduced by the plugin’s built-in data access APIs, which provide a straightforward mechanism for retrieving the results of an analysis.

The new extension is available on Redis Labs’ GitHub page alongside the numerous other tools that it has developed in recent quarters to broaden the appeal of its namesake system. Most recently, the startup released a module called RediSearch that can be used to repurpose the system into a document repository and let users quickly look up specific entries. The firm also offers a managed Redis service that has been adopted by major brands such as Twitter Inc. to support their core infrastructure.

Image via Pixabay